Production of Activated Carbon Electrode for Energy Storage Application in Supercapacitors via KOH Activation of Waste Termite Biomass

The devastating effects of termites on wood and the contribution of termite activities to the rising levels of atmospheric CO2 and CH4 constitute a serious threat to global economy and the ozone layer. In order to stall the contribution of termites to the rising levels of greenhouse gases, this work considers the conversion of termite biomass to activated carbon electrode. The waste termite biomass obtained during the production of termite biodiesel was converted to activated carbon electrode by a one-step carbonization-activation process, using potassium hydroxide as activating agent. The optimal specific surface area of the activated carbon was recorded at 900 °C, 9 h and 3:1 KOH-biomass ratio. The surface morphology, structure and composition of the activated carbon were examined using the SEM, TEM, XRD, Raman and XPS characterization techniques. The electrochemical performance of the activated carbon electrode was tested in aqueous (1 M H2SO4) and ionic liquid (1 M EMImBF4) electrolytes. Results obtained from cyclic voltammetry, galvanostatic charge–discharge and electrochemical impedance experiments showed that the specific capacitance of the activated carbon electrode was higher in 1 M H2SO4 (91.76 Fg−1 at 0.5 Ag−1) than in 1 M EMImBF4 (62.35 Fg−1 at 0.5 Ag−1). However, after completing 10, 000 chare-discharge cycles at 10 Ag−1, the activated carbon electrode lost ~ 5% of its specific capacitance in 1 M H2SO4 and ~ 2% of its capacitance in 1 M EMImBF4. Overall, the results showed that waste termite biomass could be valorised in the production of activated carbon for energy storage in supercapacitors.


Introduction
Despite the food, feed and medicinal potentials of termites [1][2][3][4][5][6][7][8][9][10][11][12], different studies have reported that termites have devastating effects on woody materials [13], contribute to the rising levels of atmospheric CO 2 and CH 4 [14] and constitute some health risks (such as allergies, microbial contamination, parasitic and chemical hazards) when consumed as food [15][16][17][18][19]. In order to promote the non-food potentials of termites and also stall the contribution of termite activities to the rising levels of CO 2 and CH 4 in the atmosphere, Brune [20] showed that the enzymatic systems of some species of termites hold great potentials for biotechnological applications for the production of new enzymes. Similarly, Koenig et al. [21], Rumpold et al. [22] and Kalu-Uka et al. [23] reported that termites represent a promising alternative for the industrial conversion of lignocellulose and lignin to biofuels, as much as they can be used to synthesize low viscous biodiesel. To further investigate the non-food potentials of termites, the present work studies the potential for the production of termite-based activated carbon electrode for energy storage application in supercapacitors.
Energy storage in supercapacitors, unlike in rechargeable batteries and fuel cells, is attracting increasing attention because of their relatively high cycling stability, high power density and fast charging-discharging capacity [24][25][26]. However, the application of supercapacitors in electric vehicles and other digital electronic devices has been limited by the low energy density associated with the mechanisms (electric double layer capacitance and the pseudocapacitance) of energy storage in supercapacitors [27][28][29][30][31]. In the past, the mechanism of energy storage in electric double layer capacitors (EDLC) was understood as an electrolyte ion adsorption process under an applied voltage. However, recent studies on the in-situ measurement of EDLC performance has revealed that the mechanism of charge storage in EDLCs was actually dependent on a three-step process of counter-ion adsorption, ion exchange and co-ion desorption in the pores of the electrode [32]. As a result, some research efforts have been made to improve the energy density of supercapacitors by either identifying new electrode materials with improved electrochemical properties or re-engineering the structural properties of existing electrode materials such as carbon nanomaterials, conducting polymers, transition metal oxides and waste biomass [28,[33][34][35][36][37][38][39][40][41][42][43][44][45].
Although the use of waste biomass (such as animal products, plants and algae) for the production of supercapacitor electrodes has been widely reported, results in the literature have also shown that the performance of biomass-derived supercapacitor electrodes does not depend only on the unique microstructure of the biomass precursors [26,30,32,33,35,46,47]. The presence of heteroatoms on the electrode surface [46], type and concentration of the electrolyte [48,49], dispersity of ion electrosorption at the electrolyte-electrode interface, and the methods of electrochemical testing, biomass carbonization and char activation have also been reported to greatly influence the performance of supercapacitor electrodes [50,51]. The attraction for the use of biomass in the production of carbon electrodes is due to their availability, natural porosity and ease of activation. Biomass can be physically activated in the presence of an oxidizing/ gasifying atmosphere at high temperature, or chemically activated using corrosive chemicals (such as KOH, ZnCl 2 and H 3 PO 4 ) as activating agents [36]. KOH activation of 1 3 biomass occurs when the KOH-biomass mixture is heated to temperatures above the melting point of KOH (360 °C). Above this temperature, the volatile materials in the biomass are vaporized while the molten KOH reacts with the carbon in the biomass to form metallic K, H 2 , and K 2 CO 3 . However, further heating of the mixture might result in the breakdown of the K 2 CO 3 into K 2 O, CO 2 and CO. By the end of the activation process, nearly all the H 2 , CO 2 and CO will be released, leaving behind a highly porous carbon material [52]. For this reason, a thorough understanding of the charge storage mechanism would depend on the optimum pore-selection of electrolyte ions, availability of passive and active pores for ion adsorption, rate of charge adsorption, and the chemical changes on the electrode surface due to ion adsorption [32].
One of the contributions of this work is the novelty of the biomass precursor, which was obtained from the waste generated during the production of biodiesel from termite biomass [23]. The termite biomass used in this study belongs to the species of Macrotermes nigeriensis. This species of termites is seasonally available in Nigeria (from March to July), and its estimated annual availability in Nigeria is 3.2 × 10 11 kg [23]. To the best of the authors' knowledge, there was no single report on the potential of termite biomass for electrochemical energy storage at the time when this work was carried out. Also, the comparison of electrochemical performance of a supercapacitor electrode in two equimolar electrolytes is not the common practice in existing literature (much less when the concentration of the electrolytes is as low as one mole). This was done in order to avoid the misgivings associated with the current practice, to reduce the risk of corrosion on the current collector due to the use of electrolytes of higher molar concentrations [49], and to promote the use of lower concentrations of environmentally toxic materials in the design of supercacitors.
M. nigeriensis activated carbon (MAC) was prepared by chemical activation method, using KOH powder as the activating agent. Different samples of the KOH-biomass mixture were activated under varying conditions of temperature, time and mixture ratio in order to optimize the specific surface area (SSA) and pore structure of MAC. The surface morphology, structure and functionalization of the optimal MAC were characterized with scanning electron microscope (SEM), transmission electron microscope (TEM), X-ray diffractometer (XRD), Raman spectroscope and X-ray photoelectron spectrometer (XPS). The supercapacitor performance of MAC electrode (MAC-E) was later examined in 1 M H 2 SO 4 and 1 M EMImBF4 electrolytes. The results showed that specific capacitance of M. nigeriensis activated carbon electrode in 1 M H 2 SO 4 ranges from 30.59 Fg −1 (at 5 Ag −1 ) to 91.76 Fg −1 at (0.5 Ag −1 ). In 1 M EMImBF4, the specific capacitance was 20 Fg −1 (at 5 Ag −1 ) and 62.35 Fg −1 at (0.5 Ag −1 ). However, at 10 Ag −1 and 10, 000 charge-discharge cycles, the activated carbon electrode lost ~ 5% of its specific capacitance in 1 M H 2 SO 4 and ~ 2% of its capacitance in 1 M EMImBF4.

Production and Characterization of MAC
A 20 g sample of the waste termite biomass was dried in an oven at 110 °C for 12 h, milled into fine particles and mixed with KOH powder, using a KOH-biomass ratio of 2:1. The mixture was activated in a tubular furnace at 500 °C for 5 h, in the presence of N 2 gas. The activated carbon which resulted from the heating process was washed with dilute 10% HCl, and repeatedly rinsed in warm water until the filtrate reached a pH value of 7 [28]. The filtrate was later dried in an oven at 120 °C for 2 h, stored in a dry glass vial, and labelled MAC500-5 h-2r. The highlights of the above production process are shown in Fig. 1.
The specific surface area (SSA) and pore structure of MAC was later optimized by repeating the above procedure using different activation temperatures, activation times, and KOH-biomass mixture ratios to produce MAC700-5 h-2r, MAC900-5 h-2r, MAC900-7 h-2r, MAC900-9 h-2r, and MAC900-9 h-3r and MAC900-9 h-4r. The optimal MAC sample was determined from results of the N 2 adsorption-desorption experiments which were performed on a BET analyzer (Micromeritics ASAP 2020 V4.02-V4.02 H). Before performing the BET analysis, the MAC samples 1 3 were degassed at 200 °C for 12 h in order to reduce the residual pressure to values less than 1.0 × 10 -3 mbar [34] and to eliminate all the surface gaseous contaminants [53]. The SSA and pore size distribution (PSD) of the MAC samples were respectively determined using the Brunauer-Emmett-Teller (BET) and the Barrett-Joyner-Halenda (BJH) methods. The total combined volume of the mesopores and macropores, V mm was calculated by subtracting the amount of N 2 gas adsorbed at a relative pressure of 0.1 from the amount of N 2 gas adsorbed at a relative pressure of 0.95 [54]. The specific surface area of the micropores, S micro was calculated from the t-plot by solving Eq. 1 [34].
where S BET is the BET surface area, and s is the slope of the t-plot.

Preparation and Electrochemical Performance of MAC-E
M. nigeriensis activated carbon electrode (MAC-E) was prepared by mixing a sample of the optimal MAC with acetylene black and binder (PVdF-HFP) in a mass ratio of 85:10:5 [28]. Afterwards, 500 mg sample of the mixture was coated on a 16 mm x 16 mm diameter nickel foil, and dried in an oven at 105 °C for 12 h to remove the organic solvent. The electrochemical performance of MAC-E was tested in an electrochemical workstation (Autolab Potentiostat/Galvanostat Model-PGSTAT 302), using MAC-E, Ag/AgCl and platinum wire as the working, reference and counter electrodes respectively. The cyclic voltammetry (CV) patterns, galvanostatic charge-discharge (GCD) profiles and electrochemical impedance spectra (EIS) of MAC-E were analyzed in 1 M H 2 SO 4 and 1 M EMImBF4 electrolytes. The CV patterns of MAC-E at different scan rates between 50 mVs −1 and 5000 mVs −1 were obtained within voltage windows of 1.0 V (for 1 M H 2 SO 4 aqueous electrolyte) and 4.0 V (for 1 M EMImBF4 ionic electrolyte) [48]. The GCD profiles were generated at different current densities between 1 and 10 Ag −1 , while the EIS readings were plotted for the frequency range from 0.01 Hz to 100 kHz. The specific capacitance, C sp Fg −1 , of MAC-E was calculated from the GCD experiments, based on Eq. 2.
where i(A) is the applied discharge current, m(g) is the active mass of the working electrode, V(V) is the voltage difference during the discharging process, and t d (s) is the discharge time.

Characterization of Waste M. nigeriensis
Waste M. nigeriensis biomass was characterized to examine the elemental composition, surface morphology, organic functional groups, crystallographic structure and thermal decomposition behaviour. Table 1 showed the elemental composition of waste M. nigeriensis biomass, based on results from both ICP-AES and CHNS analyses. Table 1 also Fig. 1 Schematic illustration of the production of activated carbon electrode from M. nigeriensis, showing the biomass precursor as waste generated from the production of termite biodiesel showed the moisture, ash, and volatile matter contents of the waste biomass, based on the proximate analysis which was performed according to the GB/T12496. 1-12,496.22(1999) standard [29]. The oxygen content of the waste biomass was however measured by the difference of C, H, N, S, and ash from 100 [55]. The results showed that waste M. nigeriensis contained relatively high oxygen, nitrogen and silicon content, which indicated that the char derived from M. nigeriensis would have high porosity, high tar content and high ash content, respectively. Figure 2a showed that the SEM image of waste M. nigeriensis was very rough and full of micrometer-sized pores. The roughness of the surface morphology could be attributed to the formation of mould and other microbial activities which occurred due to the conditions and duration of storage. Further characterization of the M. nigeriensis was performed using FTIR analysis. The FTIR plot shown in Fig. 2b was obtained after subtracting the peaks due to carbon dioxide, to avoid interference from the atmosphere [34]. Figure 2b showed broad overlapping bands at 3397 cm −1 and 3289 cm −1 which were due to stretching of the N-H amide and the alkyne C-H functional groups, respectively. The bands at 2922 cm −1 , 1656 cm −1 , and 1535 cm −1 showed aliphatic C-H stretching, aldehyde C = O stretching and C = C alkene stretching of the organic functional groups in the waste biomass. The peaks observed at 1443 cm −1 , 1393 cm −1 , 1230 cm −1 and 1077 cm −1 respectively identified some aliphatic C-H stretching, NO 2 symmetrical stretching, aromatic C-O stretching and C-O-C symmetrical stretching in the waste biomass. However, the oop bands between 1000 cm −1 and 400 cm −1 were attributed to the C-H alkyne bends due to the simultaneous deformation of multiple bonds [56].
The crystallographic structure of the waste biomass was examined from XRD analysis. The XRD image in Fig. 2c showed a broad peak at 2 = 21.5 o which signified the occurrence of amorphous cellulose [57]. Whereas the crystallographic phases between 25 o < 2 < 65 o confirmed the presence of some mineral elements in the waste biomass, the absence of broad peaks at 2 = 15 o and 34.5 o suggested that the structure of waste M. nigeriensis was void of lignin and hemicellulose, respectively [55,58].
The effect of temperature on the decomposition of waste M. nigeriensis biomass was analyzed by heating a 5 mg sample of the waste biomass in a TG-MS system from 30 °C to 1000 °C, using 10 °C/min heating rate and an inert (nitrogen) gas flow rate of 50 mL/min. The TG/DSC curve in Fig. 2d showed that ~ 85% of the total mass of the biomass was lost via a three-stage thermal decomposition process. It was observed that 3% of the total mass was lost in the first stage (30 -150 °C) due to endothermic desorption of residual water molecules and some oxygen functional groups, while 62% of the mass was lost in the second stage (250 -450 °C) due to the exothermic decomposition of residual proteins, ether extracts and fatty acids in the waste termite biomass [59]. The final stage (700 -750 °C) however showed that 20% of the total mass was lost due to the combustion reactions which resulted to ash formation [60]. The high amount of energy recorded at this stage is attributed to the heats of formation of the manganese and phosphorous oxides present in the ash. The heat released during this process could also help to refine the microstructure of the activated carbon. The fact that the DSC curve reached the baseline at 900 °C showed the waste termite biomass was completely carbonized at 900 °C [29]. However, based on the report that any biomass-derived carbon with ~ 15% volatile matter is good enough for the production of activated carbon [29], the optimal temperature range for the production of MAC was considered to be 700 -900 °C. This range of temperature is also very typical of most biomass activated carbons [28,36,41,43].   ures 3b and S3-Sb showed that MAC has a non-uniform PSD distribution, with the majority of the pores located within the range of micropores and mesopores. Details of the SSA and pore structure of all the MAC samples are shown in Table 2.

Identification of Optimal MAC Sample
Prior to KOH-activation, the SSA and average pore diameter of the waste M. nigeriensis biomass were recorded as 0.306 m 2 g −1 and 9.736 nm, respectively. However, when the biomass was activated over 5 h and the KOH-biomass ratio was 2:1, the SSA of MAC increased from 670.68 m 2 g −1 (at 500 °C) to 1064.28 m 2 g −1 (at 900 °C). The low SSA observed at 500 °C suggested that some of the pores in MAC500-5 h-2r were occupied by materials whose vaporization temperatures were above 500 °C. The maximum activation temperature was not allowed to exceed 900 °C in order to prevent the occurrence of excessive activation reactions that might affect the development of internal pores [66,67]. The table also showed that, at 900 °C and KOH-biomass ratio of 2:1, the variations in activation time had little effect on the SSA of MAC. However, when the activation temperature and time were kept at 900 °C and 9 h, the SSA of MAC increased from 1067.66 m 2 g −1 (at 2:1 KOH-biomass ratio) to 1465.56 m 2 g −1 (at 3:1 KOH-biomass ratio), and later decreased to 1440.65 m 2 g −1 (at 4:1 KOHbiomass ratio). The lower SSA observed for MAC900-9 h-2r was therefore attributed to the fact that the concentration of KOH in the mixture was not sufficient to promote the formation of the skeletal pore structure of carbon [68]. Also, the decrease in the SSA of MAC900-9 h-4r suggested that some of the excess KOH in the mixture may have formed crosslinks which resulted in the production of more volatile matter [56]. Based on these observations, the optimal MAC was identified as MAC900-9 h-3r. Although the relatively high SSA and non-uniform PSD of MAC900-9 h-3r would be required to provide the critical adsorption sites and the short transmission path required in supercapacitor electrodes [41,[69][70][71], actual performance of MAC900-9 h-3r as a supercapacitor electrode will be largely dependent on the choice of electrolyte, as well as the dynamics of the counter ion adsorption, co-ion desorption and ion exchange in the electrode [32,[72][73][74].

Surface Morphology and Structural Order
Details of the surface morphology of MAC900-9 h-3r were observed from SEM and TEM analyses, and the results are shown in Fig. 4a and b. The SEM image in Fig. 4a showed that the surface of the activated carbon was composed of a relatively ordered network of honeycomb-like tunnels. Figure 4a further confirmed that the surface porosity of waste M. nigeriensis (shown in Fig. 2a) was sufficiently enhanced during the production of MAC900-9 h-3r. The TEM image in Fig. 4b however showed that the honeycomb-like tunnel structure of MAC900-9 h-3r had fully developed micropores [29]. The structural order of MAC900-9 h-3r was examined by XRD and Raman spectroscopy, and the results are shown in Fig. 4c and d. The XRD image in Fig. 4c showed two broad peaks at 2 = 28.  (100) planes of a carbon structure [28,75]. The broadness and asymmetric nature of the (002) peak was attributed to the  low scattering angle due to the presence of γ-band associated with the amorphous structure and/or irregularity in the packing of the aromatic structure [76]. The shift in the 2 angle of the (002) plane from the 2 (= 26.6 o ) angle for the (002) plane of graphite was due to the amorphous nature and low degree of graphitization of the activated carbon [51]. The (100) peak, however, was attributed to the twodimensional reflection of X-ray from carbon layers, and the aromatic part of the carbon structure [77]. The increase in intensity of X-ray diffraction across the lower angles further suggested that the activated carbon is amorphous, and sufficiently porous [78,79]. Further details of the structural order of the activated carbon were obtained from Raman spectroscopy. The Raman spectra in Fig. 4d was analyzed within the wavelength of 100 -2000 cm −1 . The two peaks observed at 1340 cm −1 and 1588 cm −1 were respectively typical of the D band and the G-band that are found in the Raman spectra of every carbon material [28]. The broad G-band signified the stretching of the C = C sp 2 bonds in graphitic materials [28], while the D-band indicated the disorder in the graphite lattice due to the breathing mode of hexagonal rings [34,80]. The amorphous nature of MAC900-9 h-3r was confirmed by the fact that the I V ∕I G ratio is less than that of graphite [34,77] and 0.5 < I D ∕I G < 1 [81].

Surface Composition and Functionalization
The surface composition of MAC900-9 h-3r were analyzed by XPS experiments, and the results showed the presence of  Fig. 5a and b. Figure 5a showed that the major component of the C 1 s spectra was observed at 284.24 eV. This represented the sp 2 C C band for graphitic materials [77,82,83]. The peak at 286.01 eV represented the sp 3 C-C band for disordered carbon structures [84]. The band at 287.36 eV indicated the presence of C O functional group [81]. However, the O 1 s spectra in Fig. 5b showed that the C O carboxylic and -OH hydroxyl bands appeared at 527.91 eV and 531.37 eV, respectively. Although the presence of the carboxylic and hydroxyl bands suggested that MAC900-9 h-3r was not completely carbonized, their presence also predicted that the capacitance behaviour of MAC900-9 h-3r would show some pseudocapacitance effect [30,51,81,85]. The pseudocapacitance effect could however be enhanced by the amount of sulphur atoms present in MAC900-9 h-3r [86][87][88]. The slight shift from the peak positions of the carbon and oxygen species in the C 1 s and O 1 s spectra was due to the chemical nature of the neighbouring atoms on the surface of the activated carbon [34].

Effect of Scan Rate on Charge Storage
Cyclic voltametry tests were conducted to study the effect of scan rate on charge storage properties of MAC-E. Figure 6a and b showed the CV curves of MAC-E in 1 M H 2 SO 4 (aqueous) and 1 M EMImBF4 (ionic liquid) electrolytes. The CV curves were generated at varying scan rates from 50 mVs −1 and 5000 mVs −1 . The maximum voltage window for each electrolyte was determined from existing literature [33,48]. The difference in the maximum voltage windows of the CV curves is attributed to the variations in the radii of the hydrated cations and anions of the two electrolytes [89,90]. The relationship between the pores and the radius of hydration sphere ions leads to a decrease in the quantity of ions entering into the pores, and also lowers the electric double layer formation at the electrode-electrolyte interface [91]. The lower voltage window of H 2 SO 4 aqueous electrolyte could therefore be attributed to the inability of the SO 4 2− ions (which are as big as 40 molecules of water) to penetrate into some of the pores in MAC-E [91]. Although the capacitance behaviour of MAC-E was confirmed by the rectangular shape of the CV curves, the distortion in the rectangularity of the CV curves suggested that the capacitance behaviour of MAC-E was a combination of EDLC and pseudocapacitance. The pseudocapacitance behaviour was attributed to the presence of oxygen and sulphur atoms on the surface of MAC-E [51,92,93]. However, the pseudocapacitance behaviour of MAC-E was more noticeable in 1 M H 2 SO 4 than in 1 M EMImBF4. The difference in the pseudocapacitance behaviour of MAC-E in 1 M H 2 SO 4 could be attributed to the additional sulphur atoms which are present in the aqueous electrolyte. The slight suppression of the pseudocapacitance behaviour of MAC-E in 1 M EMImBF4 was attributed to the inability of the heteroatoms on the surface of MAC-E to form ionic solutions in 1 M EMImBF4 electrolyte [94].

Effect of Current Density on Capacitance Behaviour
Galvanostatic charge-discharge experiments were performed to determine the capacitance behaviour of MAC-E at different current densities, and the results for the two different electrolytes are shown in Figs. 7(a and b). The GCD curves showed a slight deviation from the normal saw-tooth shape of EDLCs which confirmed the effect of pseudocapacitance in the capacitance behaviour of MAC-E. The high voltage drop of ~ 0.3 V (in 1 M H2SO 4 ) and ~ 0.5 V (in 1 M EMImBF4) observed at the discharge section of the respective GCD curves could be attributed to the potential difference at the electrode-electrolyte interface [43].   ions, viscosity, dielectric constants and chemical stability of the two electrolytes [91]. The decrease in the specific capacitance of MAC-E as the current density increassed was however due to insufficient charge transport in and out of the micropores [34,51]. The capacitance retention of MAC-E in 1 M H 2 SO 4 and 1 M EMImBF4 was examined after 10, 000 charge-discharge cycles at 10 Ag −1 . The result in Fig. 8b showed that MAC-E retained 95% of its specific capacitance in 1 M H 2 SO 4 , and 98% of it in 1 M EMImBF4. The relatively lower capacitance retention of MAC-E in 1 M H 2 SO 4 aqueous electrolyte was attributed to the poor redox reversibility of the oxygen functional groups which decreased the electronic conductivity of MAC-E [95,96]. Although the stability of MAC-E was determined after several thousand charge-discharge cycles between the critical cell voltage and zero voltage, the authors would rather that the voltage hold (or float) testing technique was used in analysing the stability of MAC-E. This is because the charge-discharge testing technique is time-consuming and unable to correctly predict the stability of an EDLC [97][98][99][100]. Nevertheless, the results obtained from charge-discharge tests could be corrected by the fact the charge-discharge stability rating of capacitors are approximately 0.3 V higher than the stability rating obtained by the voltage float testing method [97].

Electrochemical Impedance and Capacitance Retention
Electrochemical impedance experiments were carried out to determine the electrode resistance, the electrolyte resistance and the diffusion resistance of MAC-E in 1 M H 2 SO 4 and 1 M EMImBF4 electrolytes. The values of these resistances are shown in Table 3. The disparity in the electrode resistance of MAC-E in the two different electrolytes suggested that the charge transfer resistance of MAC-E was dependent on the type of electrolyte. The Nyquist plots  for the electrochemical impedance tests are also shown in Fig. 9. At the low frequency range, the impedance spectra of MAC-E in the two electrolytes appeared as inclined lines which were shifted along the real z-axis by their respective series resistances [48]. The degree to which these lines were inclined also confirmed that the energy storage mechanism of MAC-E was a combination of EDLC and pseudocapacitance. The semicircular curves seen in the high-frequency range confirmed the porous nature of MAC-E [28,34,41]. The diameters of the two semicircular curves represented the respective resistances of the electrolytes in the electrochemical system [47,101,102].

Conclusion
This work considered the conversion of waste termite biomass into activated carbon for energy storage as a good way to combat the contribution of termite activities to the rising levels of atmospheric CO 2 and CH 4 . Activated carbon was initially produced from waste termite (Macrotermes nigeriensis) biomass through a one-step carbonization KOHactivation process. The carbonization-activation process was performed under varied conditions of temperature, time and KOH-biomass ratio in order to optimize the specific surface area and pore structure of the activated carbon. The morphology, structure and surface chemistry of the optimal activated carbon sample were analyzed using standard analytical equipment, and the results showed that termite-based activated carbon can be used for energy storage in supercapacitor. For this reason, a sample of the activated carbon was formed into an electrode via a densification process. Electrochemical experiments were later performed on the activated carbon electrode in 1 M H 2 SO 4 aqueous and 1 M EMImBF4 ionic liquid electrolytes, using a 3-electrode electrochemical system. Results of these experiments showed that the specific capacitance of the activated carbon electrode was higher in 1 M H 2 SO 4 (90.76 Fg −1 at 0.5 Ag −1 ) than in 1 M EMImBF4 (62.35 Fg −1 at 0.5 Ag −1 ). The capacitance retention of the activated carbon electrode was later examined after 10, 000 charge-discharge cycles at 10 Ag −1 ; and the result showed that activated carbon electrode retained 95% of its specific capacitance in 1 M H 2 SO 4 , and 98% of it in 1 M EMImBF4. The capacitance of the activated carbon electrode could however be improved by re-engineering the activation process, functionalizing the electrode surface, increasing the electrolyte concentration, and optimizing the thickness of the electrode, the conductivity of the current collector, and the amount/type of binder. Although the results obtained from this study have shown that termite biomass such as waste M. nigeriensis could be valorized in the production of activated carbon for energy storage, the authors were unable to determine the energy density, power density and self-discharge of the activated carbon electrode. The inability of the authors to perform the above analyses is however due to the unavailability of the equipment required for the analyses.